Severe obesity (BMIe35 kg/m2) often has early onset, is a strong risk factor for disease and early mortality, is difficult to treat short of bariatric surgery, and long-term maintenance of weight loss is usually unsuccessful. There is a significant genetic component to severe obesity, as linkage, candidate gene, and genome-wide association studies (GWAS) using single nucleotide polymorphisms (SNPs) have detected multiple genetic regions and specific genes associated with obesity. Most of the variants are common (>5%) in the population, and each explains <2% of the variance in obesity. The sum of these genes explain <10% of BMI variation, even though 40-60% of the variation is thought to be genetic. There must be significant additional genetic variation that can be uncovered by other methodologies. Severe obesity-associated regions of the genome harboring low frequency variants will be identified using both a modified GWAS haplotype approach and a pedigree-specific shared genomic sequence approach utilizing whole exome resequencing. This application proposes to overcome barriers to detecting less common or rare variants associated with severe obesity. The first barrier is that rare variants are not likely to be in high linkage disequilibrium with the common variants used in GWAS. However, we propose to infer long haplotypes of common SNPs to capture unmeasured rare variants residing on those haplotypes. Second, allelic heterogeneity in a region or gene (causal mutations occurring in different locations across the gene, e.g. MC4R for obesity or BRCA1 for breast cancer) prevents single SNPs from detecting association. The rare risk haplotypes identified across each gene or region will be collapsed into haplotype risk sets to accommodate allelic heterogeneity and increase the estimated frequency and statistical power. Third, causal variants are difficult to distinguish from the many neutral variants in unrelated cases and controls. Extended Utah pedigrees will be tested for co segregation of the identified variants in many obese relatives to weed out the neutral variants. Fourth, resequencing errors are more easily identified and removed using pedigrees by ensuring Mendelian transmission. Fifth, even pedigree-specific rare variants can be identified in extended pedigrees in which there are at least 15 meioses among 8 or more distant affected relatives. Exome capture and NextGen sequencing will identify such rare variants. Finally, a number of common variants are associated with severe obesity in our pedigrees and adjustment for these common variants will reduce the genetic variation so that the rarer variants are more easily detectable. Over 10,000 subjects, combined from various large studies, will be used for identification, replication, and characterization of identified variants. The highly informative set of pedigrees, cases/controls, and population-based cohorts will be used to further unravel the complexity of the genetic underpinnings of severe obesity and should lead to the identification of rare variants and an understanding of the physiological pathways through which severe obesity develops, suggesting ways to prevent or reduce severe obesity.